Disadvantage for the incumbent or not?

There is no correlation between voter turnout and incumbency

Conventional wisdom has always been that a higher voter turnout corresponds with a greater anti-incumbency effect in India. One of the many reasons noted in the loss of the UPA in the 2014 elections was the presence of a higher voter turnout in comparison to 2009. But in the last decade and a half, more incumbent governments have been re-elected even as voter turnout has risen. Is there any correlation between voter turnout and incumbency?

In the U.S., there is a clear negative association between incumbency and turnout. One explanation for this is the competition effect — greater the competition in a bipolar election, higher the participation, which will include more dissatisfied voters. While the hypothesis of high partisanship in a bipolar political party system holds true for the U.S., it does not for India.

Milan Vaishnav and Johnathan Guy in a paper titled ‘Does Higher Turnout Hurt Incumbents?’, published in the journal

Studies in Indian Politics (April 2018), look at Assembly elections in 18 major States between 1980 and 2012 to answer the question of correlation. They find that an increase in turnout has no statistically meaningful relationship with incumbent performance — either in the re-election of the incumbent or in the change in vote and seat shares from the previous election for the party.

A cursory look at the overall relationship between voter turnout and the three variables — re-election, changes in vote share and seat share — for the 128 elections held in that period shows that this is skewed towards anti-incumbency due to outlier elections in States such as J&K and Punjab which went through periods of insurgency. The authors omit these State elections (as outliers) and then study the same relationship. They find that there is no clear pattern. They look at political contexts when anti-incumbency is related to voter turnout and find no correlation there either.

What is interesting is that State elections in India have tended to return more incumbents since the 2000s (42.6%), a number closer to the trend in the 1980s (44.7%), which fell to 21.1% in the 1990s. The nineties was a period of significant political turmoil and greater regionalism and federalisation of the Indian polity. The authors suggest looking at structural changes in voting behaviour that seem to support greater incumbency rather than looking at incorrect and simplistic assumptions about higher turnout indicating greater anti-incumbency.